The purpose of this prospective study is to identify risk factors for transition into injection drug use among young adult non-injection drug users (heroin, crack/cocaine users) in New York City. Extensive data has shown that injection drug use is a substantially higher drug-using health risk practice than non-injection primarily due to transmission of blood-borne infectious disease, psychiatric co-morbidity, and other social consequences. Risk factors for illicit drug use have been studied; however, data specifically on transition into the practice of injection remain sparse. Researchers now suggest that determining the effects of the social environment such as social networks and social support on transition into injection as the next step in advancing our knowledge of both injection drug use and injection-related health consequences. For example, while research has shown whites (and according to some studies, Hispanics) as being more likely to use illicit drugs, and to begin use at a younger age than their black counterparts, black and Hispanic drug users are more likely to burden higher rates of infectious disease compared to white drug users. These racial/ethnic differences support the study of social context and how factors that characterize this context may change over time (i.e. from pre-transition to transition). Research has indicated social networks as a viable method of disseminating health-related information lending to additional public health impact resulting from this line of research. Thus, providing a better understanding of ones' social network/support can lead to public health strategies that target social networks. In this proposed study, we will determine the effect of social networks/social support characteristics (e.g. race/ethnicity of index and network members, drug use norms, network size/composition, and economic, informational and emotional support) on: 1) transition into injection, 2) age at transition, and 3) early sex/drug-risk behavior following transition. Data will be obtained from: (1) a prospective study with 2 years of follow-up among 800 non-IDUs, aged 18-35; and (2) a cross-sectional survey among 400 recent onset IDUs, injecting <=2 years, aged 18-35. The study involves targeted and respondent driven sampling methodology of drug users across NYC. The baseline and cross-sectional surveys will include a four-year sociobehavioral review that will capture 2-3 years prior to transition into injection among IDUs that can be compared to the corresponding calendar period in non-IDUs. This study will include analyses for case-control (new IDU cases/non-IDU controls), cross-sectional (new IDUs), and prospective (non-IDUs) designs. Logistic regression analyses and marginal models will be constructed using generalized estimating equations to account for potential intra-class correlation of repeated measures within an individual over time.